Ground Balance

[Carl-NC]

The idea I had was to get the output of the GB/EF2 (being DC level) and inverting it with gain (I might be getting mixed up with inverting the GB, maybe it's got to be non inverting). Then the output of this stage would be fed to a variable resistor and on the other side have the non inverting BG/EF1. This would hopefully work as the ground balance. Do you think this would work or would you have any better suggestions.

This should work. TDI applies the GB to a higher gain stage, then subtracts GB from the main channel in a subtraction diffamp. Not necessarily better, just different.

[Mick-GD]

Hi Carl, I'm just after a little more guidance. So far the design is the same as the circuit I posted earlier I have changed the 10k EF pots to 500ohms and have both the Target and GB samples balanced into two separate integrators and the outputs showing a DC level depending on Target distance. Both sample outputs go in a positive direction measured from common (9 volts in my circuit). The GB channel is a lot lower in level because it's being sampled out further.

The idea I had was to get the output of the GB/EF2 (being DC level) and inverting it with gain (I might be getting mixed up with inverting the GB, maybe it's got to be non inverting). Then the output of this stage would be fed to a variable resistor and on the other side have the non inverting BG/EF1. This would hopefully work as the ground balance. Do you think this would work or would you have any better suggestions.


Cheers
Mick


Here's a couple of pictures for the detector build so far. If the design works correctly the final version will go onto a properly designed board.

[Mick-GD]

Hi Mick,

I got bored and made this spreadsheet, its a sim of the TDI method, you might find it useful. You can play around with the sample points, co-efficients etc. and see how it affects the response to different TC targets. Not sure how valid it all is but hey its a bit of fun.

Midas

Hi Midas, thanks that might come in handy depending on if I can get the design to work.



Mick

[Midas]

Hi Mick,

I got bored and made this spreadsheet, its a sim of the TDI method, you might find it useful. You can play around with the sample points, co-efficients etc. and see how it affects the response to different TC targets. Not sure how valid it all is but hey its a bit of fun.

Midas

[Mick-GD]

TDI EF=75us. Long TC targets are still detected in the EF sample, but should always be substantially weaker than the main sample. The result is it doesn't matter much.

Hi Carl thanks for the reply.

At the moment the circuit is only on one of those white plugin prototype boards but I'm transferring it to a breadboard design at the moment. I also have to make a better +/-5 volt power supply with the circuit running at 9 volts input I can't sync the 7660 to the TX pulses if I do for the output into the 5 volt reg is too low.



Mick

[Carl-NC]

TDI EF=75us. Long TC targets are still detected in the EF sample, but should always be substantially weaker than the main sample. The result is it doesn't matter much.

[Mick-GD]

Hi Carl, back in post 24 I wrote that when looking at the EF1 sample separately and the delay set to about 200us larger targets lead and steel still get detected at a much lower level. I tried at 300us and the lead and steel targets still get detected but once the EF1 sample was set to 500us the lead is no longer detected but the steel target is still detected.


What I'm wanting to know is does the TDI have it's EF sample points out this far and or does it really matter if a target is being detected in the EF samples ?



Cheers
Mick

[Orbit]

That , I used a brick of 400gr ,of course with f2 channel you will get what expected ! a thank you Carl about clarification solutions about VCO and Threshold na TDI ,thank you once more Karll ...

[Mick-GD]

Hello Mick ,I also tried something like with magnet
really function , but I tried to brick red with much particle iron ,and will not it has response .I think talk only about one channel ,if I understand your ,,

Hi Orbit, the magnets worked well I was able to null out the EF (earth field) signal at the output of the integrator.

With the red brick test I could see an output being produced when the brick was waved over the coil I should also mention the red brick produced less responce than a one gram lead weight at the same distance, Once I have the GB and EF2 samples set up It should adjust out easily. Fingers crossed.



Mick

[Orbit]

provided you follow stability of output whether there chatter

[Orbit]

Hello Mick ,I also tried something like with magnet
really function , but I tried to brick red with much particle iron ,and will not it has response .I think talk only about one channel ,if I understand your ,,

[Mick-GD]

Usually the signal has been amplified to the point there isn't enough headroom for any more gain before the integrator. So the add'l gain is applied after the integrator. Or, you can apply the extra gain within the integrator itself.

I actually had the frontend rail today with a big target so I'm going to reduce the frontend gain a little and add the gain after the integrator.

Something else I noticed today when working on the target and EF1 sample circuit at one stage I had the circuit setup with only the EF1 sample connected directly to the integrator and notice it still picked up some bigger targets one was a big screwdriver and the other was a five ounce lead weight. I'm guessing we wan't to minimize target response in the EF samples because anything thats out that far should be rubbish only. The EF1 sample is set at about 200us at the moment so i'll adjust it out to 300 or 350 tomorrow and see what happens.

I should also mention at the moment I'm still running 100us TX and a receive time of 450us till next TX.



Cheers
Mick-DG

[Carl-NC]

Carl, with regards to the output of both the integrator's would it be best if I amplifiy the GB pre the GB/EF2 integrator or after when it becomes a DC value of both GB/EF2 ?

Usually the signal has been amplified to the point there isn't enough headroom for any more gain before the integrator. So the add'l gain is applied after the integrator. Or, you can apply the extra gain within the integrator itself.

[Mick-GD]

Ok I found a few different magnets and all showed no response using the same circuit.

I then decided to have another look at the balance circuit. I changed the 10k resistor back to 3.9k on the inverting amp and changed the pot to a 500 ohm that I had tried earlier and I was also able to balance the EF1 out. All I can think of is maybe I had a wiring issue when I tried the 500 ohm pot the first time.

Once I have the GB and EF2 circuit built I'll post up a picture of the modified circuit.


Carl, with regards to the output of both the integrator's would it be best if I amplifiy the GB pre the GB/EF2 integrator or after when it becomes a DC value of both GB/EF2 ?

[Mick-GD]

No worries I'll go find a few different magnets and do some more testing and report back shortly.